End wind gave no clear first-block bird richness link in four Canada and US samples in 1985-2024

We tested whether end wind predicted recorded bird taxa in the first ten survey stops after control for start weather. Four models used 64,681 accepted Canada and US surveys in 1985-2004 and 2005-2024. All four first-block intervals included zero after correction. First-block slopes ranged from -0.091 to -0.005 taxa per higher wind code. Last-block slopes ranged from -0.913 to -0.604. Last-minus-first slopes were negative in all samples. This pattern is consistent with a more negative association later in a route. It does not prove a detection mechanism. Wind persists, end wind can mark early conditions, and blocks differ in place and time. Selection, exposure error and shared routes limit inference. We reused a verified published panel and did not rebuild raw data. The package includes the panel, code, provenance, all fits and a figure. We did not test climate trends, wind damage or pollution.

birds · observation bias · wind · temporal specificity · recorded taxa · Breeding Bird Survey